Project Summary/Abstract Tremendous advances in the molecular analysis of melanomas and their immune microenvironment have provided mechanistic insights, which in turn have translated into remarkable improvements to treat and even cure patients with advanced disease. However, many patients still succumb to their disease, in part because treatment starts too late. Identifying patients in need of systemic treatment earlier and removing pre-neoplastic lesions before they evolve into cancer are promising effective ways to reduce melanoma-related deaths. Current screening methods can detect lesions with a broad spectrum of natural behavior-- from those with lethal potential to the completely benign. There is an urgent need to develop precision molecular tests to: (1) Distinguish between benign and malignant lesions (i.e., diagnostic biomarkers); (2) Differentiate biologically indolent versus aggressive tumors in need of additional treatment (i.e., prognostic biomarkers); (3) Identify high-risk individuals who should be screened (i.e., screening biomarkers). My group is performing multi-omics studies on primary lesions to study the earliest molecular and cellular events associated with melanoma initiation and evolution to obtain key insights that are essential to improve risk stratification of early stage cancers, cancer prevention and detection. Characterizing the genetic evolution of main melanoma subtypes as they develop from their respective pre-neoplastic lesions is one major goal of this proposal, as is developing a biomarker-based taxonomy, which will serve as framework for developing significantly advanced diagnostic and prognostic algorithms. Studies of the tumor genome will be complemented by characterization of the composition and functional state of immune cell infiltrates during melanoma evolution to shed light on the interaction between the immune cell infiltrate and tumor cells. The combined analysis of tumor genomes and host response will reveal key mechanisms driving elimination, equilibrium, and ultimate escape and allow us to extract biomarkers, which will undergo validation in patients with known outcomes with the goal to develop clinically sound tests. Specifically, we will formulate candidate algorithms that can (i) better anticipate metastatic dissemination than current methods and (ii) predict whether a pre-neoplastic lesion is likely to progress to melanoma, and validate them using patient cohorts with known follow-up information.